Edge detection method, device and equipment based on predicted clear boundary

Abstract

The invention discloses an edge detection method, device and equipment based on a predicted clear boundary, and the method comprises the steps: constructing an initial deep convolutional neural network, and carrying out the training of the deep convolutional neural network to generate a target network; obtaining a to-be-identified image, and inputting the to-be-identified image into the target network; obtaining an output result of the target network, and obtaining a predicted boundary according to the output result of the target network; and detecting the edge according to the predicted boundary. According to the embodiment of the invention, rich hierarchical features can be automatically learned, fuzziness in prediction is eliminated, and a clear prediction result can be predicted without post-processing. According to the network, hierarchical features are effectively utilized, a boundary mask with high pixel precision is generated, and convenience is provided for edge detection.

Description

technical field [0001] The present invention relates to the technical field of image processing, in particular to an edge detection method, device and equipment based on predicting clear boundaries. Background technique [0002] Edge detection is a long-standing task in computer vision. In the early days, the goal was defined as finding discontinuous sudden changes in the intensity image. Currently, localization of semantically meaningful object boundaries is expected to play a fundamental and important role in computer vision tasks such as image segmentation and optical flow. [0003] In the past few years, deep convolutional neural networks have dominated the research of edge detection. While methods based on deep convolutional neural networks are good at generating semantically meaningful contours, we observe a general phenomenon that their predictions are much thicker than those of classical methods. One possible reason for the prediction boundary being too thick is t...

AI Technical Summary

Problems solved by technology

[0005] In view of the above deficiencies in the prior art, the object of the present invention is to provide an edge detection method, device and equipment based on predicting a clear boundary, aiming to solve the problem of poor edge detection quality due to the thick boundary predicted by the edge detection method in the prior art , technical problems with low accuracy

Images